A typical multiple-input-multiple-output (MIMO) network comprises a base transceiver station (BTS or BS) with multiple antennas and multiple mobile stations (MS), at least one of which has multiple antennas. The performance of an MIMO network can be improved by employing a beamforming technique.
A base transceiver station and a mobile station in an MIMO network employing a beamforming technique communicate with each other with beamformed signals. The BS computes beamforming weighting vectors using receiving signals that are coherently combined. The signals transmitted from the multiple antennas on the BS are weighted based on phase and magnitude so that they will be coherently combined at the receiving MS.
Assuming that there are M antennas on the BS and N antennas on one of the MSs, there will be M×N MIMO channels between the BS and the MS. By applying beamforming weighting vectors to the multiple antennas on the BS, L beamformed channels are created between the BS and the MS, where L<=min(M, N). The quality of the beamforming weighting vector is essential to the performance of the L beamformed channels. An optimal beamforming weighting vector maximizes beamforming gain and at the same time achieves the de-correlation of signals on the L beamformed channels so that the receiver of the MS can easily detect MIMO signals.
Several methods have been developed to compute beamforming weighting vectors for the BS using the signals transmitted from the multiple antennas on the MS. The beamforming weighting vectors are applied to the multiple antennas on the BS to de-correlate the transmitting signals on the beamformed channels.
One frequently used method for computing beamforming weighting vectors is to find the primary eigenvector of a covariance eigenvalue problem describing the communication channel. In this method, the signals transmitted from the target antenna are treated as the desired signals while those transmitted from the remaining antennas are treated as interference signals. By partially nulling out the interference signals, the transmitting signals on the beamformed channels are de-correlated
As such what is desired is a method and system for creating beamformed MIMO channels with partial nulling.